QuickSet for Voice Assistants & Applications

Smart homes today are rapidly evolving with many interesting use cases of artificial intelligence making way into our homes. These AI implementations are mainly done through smart speakers or voice remotes which lack the fundamental properties of  (1) compatibility with mainstream services and devices that consumers care about, referred to as “skills,” and (2) a basic understanding of context which includes nearby "things" and their current state.

A basic understanding of context

A true knowledge graph of devices and capabilities can apply context to user commands, and enable compatibility

LG Electronics’ has transformed their smartphones into a whole-home personal assistant which illustrates the vision of handheld device manufacturer in offering a whole home solution to their consumers.  Although the smart speaker category has grown rapidly to provide the experience of a personal assistant within the home, even the most popular solutions lack some of the fundamental features. An example of such a shortcoming is the daily content consumption of common AV systems in households because of the lack of understanding for these devices. These solutions must be able to identify the devices, their associated properties and content, and easily switch between them to provide what is called as a smooth user experience. A true knowledge graph will apply context to the user commands. At the end, when a consumer says "Launch Netflix", it should properly decipher this intent, find this app on a preferred device and launch it.

A true home Buttler should provide a consistent and hassle-free experience in discovering and interacting with nearby devices at home agnostic of the underlying communication protocol. This multi-protocol requirement, as well as context knowledge, are core capabilities of QuickSet and they represent the true vision of a connected home. In this vision, infrared and RF compatibility is as essential as IP capabilities, all of which is offered in the latest release of QuickSet.

 

Device status discovery addresses the challenge of limited contextual understanding of current voice assistant systems. A consumer should be able to use a simple command such as “pause” or “what’s playing?” instead of an unnatural and verbose version where they are trying to convey the context. In real-life scenarios, content playback can be initiated from different control points, or through different playback devices. Media watermarking and recognition techniques may be employed but can be limiting in reliability, cost and efficiency.  A hybrid implementation which includes a dynamic and device-centric discovery platform would enable a scalable and reliable solution. Alternative implementations fail to deliver this when they meet the realities of a consumer’s home where multiple brands and providers need to work together in a single environment. So if a voice assistant of the smart home is will be the future, it must address the above-mentioned realities of consumer home.

Device status discovery addresses the limited contextual
understanding of voice assistant systems.